Re: [Patents] The Importance of Potential (continued)

[Hartmut Pilch <phm@a2e.de>]

> I have actually read the document of the German patent court before. But
> while I agree with their instincts, I did not see any concrete legal
> argument as to why software wouldn't be patentable under their findings.
> They also did not set realistic restrictions.

They set restrictions which are applicable for the Dispositionsprogramm 
application just as for lots of other applications and which you can apply
today on test samples of patents.

I'm afraid you haven't tried it and don't understand it.

This is all the more worrisome as you are entering a debate which has 
already reached a certain maturity and where the Parliament has already 
adopted the doctrine which you are failing to appreciate.
 
> "According to this, a patentable invention is a teaching for plan-conformant
> action utilising controllable forces of nature for achieving a causally
> overseeable result. It is beyond doubt that a rule of organisation and
> calculation, as described in this patent application, constitutes an
> instruction for plan-conformant action and that the execution of this
> instruction leads to a causally overseeable result. "
 
> As we can see the only differentiation they could make was that a
> mathematical rule did not contain the "forces of nature" part in itself.  
> Are electrons not controlled forces of nature? The answer is that all
> technology by definition is a controllable force of nature.

Yes, but a program consists only of symbolic entities, not of electrons.
If you dress up your claim to include non-inventive subject matter such as
a diskette or "technical effects" on a computer, then you have forces of
nature in the claim, but not in the solution which is rewarded by a 
patent.
 
> "However human mental activity does not belong to the controllable forces of
> nature in the sense explained above. This concept refers only to those
> forces of nature which lie outside the activity of the human mind and are
> controlled by man with help of the human mind. "
> 
> Does this necessarily mean a moving apparatus, how would this include a
> washing machine and exclude a computer apparatus or a computer network?
> Doesn't a computer operate outside the mind?

As said before, they must be part of the allegedly novel solution which is
to be rewarded by a monopoly.  I.e. part of the broadest claim (with the
least features)  that still contains the whole solution.  The computer
apparatus is usually not a part of this solution.  German examiners/courts
would say "the invention-relevant solution is completed within a virtual
machine before, during its non-inventive realisation on a knwon computer
the field of technology is set foot upon".

The "forces of nature" doctrine, like any definition of what is an 
invention, including the negative definition of Art 52 EPC, works only if 
you apply it to the achievement in a narrow sense (i.e. the broadest 
possible claim that contains the achievement) and not to just any kind of
claim with extension features which the applicant is free to insert.
 
> Although the court had prevented patenting of the human mind which is
> admirable (it's probably up for grabs now) , they had not given a realistic
> definition of what could constitute a "force of nature". 

The four forces of nature are found in every physics textbook.

> In regarding the
> question as to what could be defined as a "force of nature" they came up
> with the inspiring idea,
 
> "In the Rote Taube (red dove) decision, this court generalised this
> definition so as to accommodate other forces of nature than those of physics
> and chemistry, e.g. those of biology. "
 
> Great... so we have a definition of an invention! One that must include
> biology or at least a realm outside of physics and chemistry. What possible
> clarity can this give when we talk to politicians about how we decide what
> is patentable or not. Tell them that it must contain "forces of nature other
> than those of physics and chemistry,?"
 
You mean this gives us problems when talking to opponents of "life 
patents" ?

I don't think so.  The court didn't say how far extensions of the concept 
of "forces of nature" should be imperative.  It just said that there is 
room for adjustment as science progresses.  The concept is flexible and 
adaptable to progress of science, but at the same time it has clear 
limits.  It does not allow the patenting of abstract or 
non-deterministic solutions. 
 
> 'no "considerations located in the technical realm" '
> 
> I like this quote because we all know what they were trying to get at.
> However how do we apply this in reality to a washing method on a washing
> machine and not to a digital decoding method for a mobile phone, or a
> digital method for controlling a computer apparatus? They all use
> controllable forces of nature, none of them use biology except perhaps when
> they are operated..
 
As said before: you don't just apply this to anything the applicant choses 
to formulate but to what he actually contributed, the 
achievement/solution/teaching/contribution which qualifies him for 
remuneration. 
 
> "it must serve to achieve the causally overseeable result",
> 
> What does this mean? 

This excludes non-deterministic processes such as ones based in esthetics 
or social science.

> Does a monitor that behaves under the influence of software not have a
> causally overseeable result? Does a computer that generates graphics not
> have a causally overseeable result?

Of course it does.
But the solution is completed before its application to a known computer.
There is nothing inventive in knowing that a the Von Neumann model can
be applied to hardware. 
 
> I must say that you place a lot of emphasis on obsolete case law. 

It is not obsolete.  
It is used by some courts and was reconfirmed by the European Parliament.
 
> It is obsolete for a reason. These cases did not answer the question in
> reality of what was patentable.

EPO wants it to be obsolete for a reason:

 (1) It answers the questions in a painfully clear way which conflicts 
     with their policy goals
 (2) There are enough people who don't get it, which makes it easy to
     create confusion about it and discredit it. 

> They only showed admirable instincts as to why mathematical methods
> should not be by talking generally about mind and matter. But whether
> these methods improve the efficiency of a network or spin a washing
> machine in a new way is not the issue, only the EPO have made it one.

Effects are not the issue at all.
The issue is whether the solution is completed within a symbolic 
construction such as a program.
 
> When we get into a situation that we ourselves are arguing that software by
> itself can render a new patentable process when implemented by an existing
> machine then why are we complaining about "computer implemented inventions"
> at all? 

Nobody said that software renders the washing process patentable or that
it is implemented by software.  Software is nothing, just an abstraction.
Nothing material can be implemented by software.

> The answer is because that method by necessity involves the expression
> of the software.

I don't quite understand, but any method arguably involves the expression 
of software.  You can always abstract from a process and describe it in a 
formal language.
 
>  e.g. Transmittion of data over a network. (The network is physical and
> therefore "technical" and can be made to produce a patentable method that
> uses "forces of nature" but it is not the network but a software program
> that does this ).

As said before, whether the claimed objects are physical is irrelevant.
You must know what you apply the invention test to. 
 
> My proposal was not to "re-invent the wheel", rather to cut through all the
> smoke that has been generated by the EPO since 1978 and to find a coherant,
> simple policy on this that is both philosophically and legally coherant and
> that can be explained to the public and politicians simply.

We already have that, and it was approved by the European Parliament.
So if you want to improve on it, you will harm us unless you fully 
appreciate what is already there.
 
> The argument that I want to get across has to be routed in law. (not the
> EPO's or patent courts  version of it). 

As is the case with what we already have.
And additional rooting in caselaw is not a disadvantage.

> It is currently as follows:
 
> 1) Computer programs are synonymous with the "technical effect" or
> "technical process" they produce (as demonstrated by the dictionary
> definition of a "computer program" and the software copyright directive
> 1991).
> 
> 2) This technical effect (the execution) is covered "as a literary work" 
> by copyright and is therefore by itself unpatentable.

Good so far.
 
> 3)Any invention therefore that only has novelty by means of a "software
> innovation" running on an existing apparatus by itself is not an invention
> whether it displays a technical improvement or not because that process is
> synonymous with software.

As shown by the "automated chemistry lab" example, "existing apparatus"  
can not be the criterion.

Is the chemical process then the "further technical effect" about which 
the EPO was talking ?

> Anything less than this and we may as well hand the argument over to the
> patent lobby because it will not hold together.

Make it as simple as possible but not simpler.

My impression is that a lot of people are unable to understand the
perfectly working tests that are already there and are looking for simpler
ones that do not hold together.

> The other alternative is ,
> 
> "while software per se is not patentable, if by means of that software a
> method is realised whereby that method relates to industrial application, is
> new then it should be patentable, but not the software itself"
> 
> When asked how we decide this we would have to talk to them about
> "controllable forces of nature" that operate above and beyond "physics and
> chemistry". I can see the patent lobby laughing now...

You are creating the basis for their laughing and I'm sure your text will
be quoted.

You can solve the problem only if you differentiate between what is
claimed and what was achieved and apply whatever test you have to the
latter.  The difficult part is the differentiation, the isolation of the
"invention", not the test of whether something is an invention.  Once you
have that, you can either use Art 52 or the "forces of nature" test
interchangeably.  Without it you will not be able to use any test.
 
>  If we want to stop pure software patenting, we have to draw a line in the
> sand somewhere and a process relying solely on software is a good place to
> start . Otherwise any process implemented by an "apparatus" of some
> description running under software that has an application to an industry of
> some description is patentable and software would by necessity constitute
> the "invention".

If software constitutes the invention, then the process is not patentable 
under Art 52 EPC as well as under the "forces of nature" doctrine.

Your problem is the methodology to find out what constitutes the
invention.  Software does not constitute the invention just by because new
software is run on known equipment.

Btw we already have the lines drawn in the sand.  What you are doing now 
is mainly trying to deconstruct what we have and what the Parliament has 
approved and to replace it with something new that doesn't hold, because
of flaws which have already been discussed in the patent commentaries at 
the time when the solutions which the Parliament is proposing were tested 
and proven in practise.

-- 
Hartmut Pilch, FFII & Eurolinux Alliance              tel. +49-89-18979927
Protecting Innovation against Patent Inflation       http://swpat.ffii.org/
300,000 votes 2000 firms against software patents    http://noepatents.org/